Detect CVE-2026-54769: Langroid TableChatAgent Sandbox Escape via eval() RCE in Elastic Security
Detects exploitation of CVE-2026-54769, a critical sandbox escape vulnerability in Langroid's TableChatAgent component (versions <= 0.65.1). The incomplete eval() mitigation allows attackers to craft malicious inputs that escape the intended sandbox and achieve remote code execution on the host system. CVSS 10.0 with public PoC available.
MITRE ATT&CK
- Tactic
- Execution Privilege Escalation
Elastic Detection Query
sequence by host.name with maxspan=5m
[
network where event.category == "network" and
network.direction == "ingress" and
(
url.path : ("*chat*", "*agent*", "*query*", "*tablechat*") and
(
url.query : ("*__import__*", "*__builtins__*", "*os.system*", "*subprocess*", "*__subclasses__*", "*__mro__*") or
http.request.body.content : ("*__import__*", "*__builtins__*", "*os.system*", "*subprocess*", "*__subclasses__*", "*__mro__*")
)
)
] by source.ip
[
process where event.type == "start" and
process.parent.name : ("python", "python3", "uvicorn", "gunicorn") and
(
process.name : ("bash", "sh", "dash", "curl", "wget", "nc", "ncat", "python", "python3") or
process.args : ("*-c*", "*exec*", "*import*", "*/bin/sh*")
)
] by host.name
| head 100EQL sequence detection correlating inbound HTTP requests with eval() sandbox escape payloads to Langroid endpoints, followed within 5 minutes by suspicious child process spawning from Python web server processes — indicating successful RCE.
Data Sources
Required Tables
False Positives & Tuning
- Authorized red team exercises exploiting CVE-2026-54769 in test environments
- Python-based automation frameworks spawning subprocesses for legitimate data processing tasks
- Development pipelines running Python scripts via web-triggered hooks on the same host as Langroid
- Monitoring agents that spawn shell processes from Python parent processes for health checks
Other platforms for CVE-2026-54769
Testing Methodology
Validate this detection against 4 adversary techniques from Atomic Red Team. Each test below lists the behaviour to exercise and the telemetry you should expect to see. Executable commands and cleanup steps are available with Pro.
- Test 1Langroid TableChatAgent eval() Basic Sandbox Escape via __import__
Expected signal: Python process executing os.system() call; creation of /tmp/cve_2026_54769_poc.txt by the Python process; Syslog/auditd records showing system() syscall from python3 process
- Test 2Langroid TableChatAgent MRO Traversal Sandbox Escape for RCE
Expected signal: Python process spawning subprocess.Popen with system commands; process tree showing python3 as parent of id/sh commands; auditd EXECVE records for child process
- Test 3CVE-2026-54769 HTTP API Exploitation Simulation with Reverse Shell Payload
Expected signal: Outbound TCP connection from Python/uvicorn process to attacker IP:4444; bash process spawned as child of python3/uvicorn; network flow showing new connection to port 4444 from the Langroid server IP; HTTP access log entry with encoded payload in POST body
- Test 4Langroid Package Version Enumeration for CVE-2026-54769 Exposure Assessment
Expected signal: pip command execution; find/grep commands scanning for requirements files; curl requests to application health/version endpoints; process enumeration via ps
Unlock Pro Content
Get the full detection package for CVE-2026-54769 including response playbook, investigation guide, and atomic red team tests.